A Novel 3D Reconstruction Algorithm of Motion-Blurred CT Image.

Area of Science:

• Medical Imaging
• Computer Vision
• Algorithm Development

Background:

• Patient motion during CT scans causes image blurring, hindering realistic 3D reconstruction.
• Accurate 3D visualization is crucial for medical diagnosis and treatment planning.

Purpose of the Study:

• To develop an effective deblurring method for motion-blurred CT images.
• To propose an advanced 3D reconstruction algorithm for improved accuracy.

Main Methods:

• A Generative Adversarial Network (GAN) image translation model was employed for deblurring.
• A novel Marching Cubes (MC) algorithm, GI-MC, integrating golden section and isosurface smoothing, was developed for 3D reconstruction.

Main Results:

• The GAN deblurring algorithm demonstrated superior restoration compared to existing methods, evaluated by Shannon entropy ratio and peak signal-to-noise ratio.
• The GI-MC algorithm achieved higher reconstruction accuracy for liver patients, outperforming traditional MC, Li's, and Pratomo's algorithms by 9.9%, 7.7%, and 3.9%, respectively.

Conclusions:

• The proposed GAN deblurring and GI-MC reconstruction methods significantly improve the quality and accuracy of CT 3D visualizations.
• These advancements offer more realistic and reliable imaging solutions for medical professionals.